52/52 semi-integrated routing headset, how hard can it be?

[adam900710]

Hi,

Although I hate full integrated cable routing (aka, cable/hose through the headset bearing), I'm afraid there will no longer be any light frame that allows cable/hose to go through the head tube/down tube.

Thus I'm tinkering with my spare Bigrock frame about some semi-integrated cable routing (aka, still go through the upper headset bearing, but not directly into the barstem).

First of all, I'm not asking for things like mechanical shifting. Just like everyone else, I go wireless Di2, so only two brake hoses and this should be a pretty low bar to clear.
But I'm still very surprised by how bad the design can be, especially from FSA.

So far I haven't found any good alternatives, the ones I tried are:

• The bigrock one, FSA-like type

AKA, the one with a cut out at the front of the compression ring, just like FSA.

And I call it BS.

The cut out makes the contact area super small, thus will easily crush the carbon steering tube.

• The bigrock one, star compression ring type

The design is better, but won't work with the provided headset cover.
The compression ring doesn't seem to have enough space for brake hoses.

So again garbage.

• Shimano/Pro fully integrated cable routing headset

The design is much better, but I still don't like the compression ring spacing.

The problem is, hoses can only go through the two openings at the front (11 and 1 o'clock).
The other openings at 10 and 2 o'clock are not large enough for hoses, only for shifter cables.

And that's why the Shimano one is only allowing cable routing just under the stem, and put a lot of pressure on the hoses.

• TOKEN S-BOX

Originally I have the highest expectation on this one, as I think I can finally put hoses through 10/2 o'clock openings.
But nope, the compression ring is the same one as the Shimano headset.
So still the same fixed 11/1 o'clock for hoses.

• Neco H332MP-1WR

Finally I can put my hoses at 9/3 o'clock, but more problems are coming and eventually makes it unusable.

Firstly, the depth of the upper bearing seat must be very shallow to make it work. The official drawing requires a 7mm depth.
But my Bigrock frame has a depth around 8.5mm, meaning either play or super stiff steering.

It can be solved by ordering a 42ID 52OD, 45/45degree 8mm depth bearing to replace the upper one though.

Secondly the seal around the hoses are pretty bad, just some cuts on the rubber, and you call it a water proof seal?

• ACROS 52/52 ICR

Just ordered, still on the way.

But so far it ticks all the boxes, proper compression ring design, 9/3 o'lock cable/hose entrance, proper water proof seal around the cable/hose entrance.
[/list]

I'm surprised how hard it is to find a properly designed 52/52 headset.

I'd say the Neco one is near perfect, just need more bearing choices to fit more frames, and a better headset cover design.
The compression ring design of Shimano/TOKEN is good, but fixed hose at 11/1 o'clock is not good enough.
Although the TOKEN S-box has a pretty good headset cover to seal the cable entrances.

The biggest disappointment is, no Chinese headset is doing anything sane, but blindly follows the cursed FSA design.
Only Taiwanese Neco seems to know what they are doing, and no one even bothers to copy ACROS.

[Sakizashi]

Hey, so I spend a lot of time with headset bearing since I run a microbusiness that helps people jailbreak from proprietary systems.

There are many misconceptions in this post. The most important consideration in headset design is how well it enables the bike to hold preload. The designs that are open in front are often better than those with routing on the sides for this. The contact area on the front is a minor part of the overall load path in these designs and is often offset by having a deeper C ring (see the FSA H2630A c ring or the Deda designs for example).

The other thing to think about is how the hoses run. With the gap in front, the hoses can easily run down without needing extra stack height that takes them to the side. Different manufacturers handle this tradeoff differently. Pinarello is an extreme example where the routing path complexity is handled in the stem. Most do it like Specialized, Enve, and Acros, and do it in the spacer stack.

I don't argue with the premise that there is a lot of room for improvement, particularly with designs that are only made for 2 hoses, but I regretted playing around with the Acros system; it's unnecessarily complex and a PITA to get replacement parts for. The Acros design actually has an internal spacer, then a traditional metal compression ring, and then a headset cover. FSA, Token, Enve, Deda all use 2 pieces, just a headset cover and a compression ring. Compatible HT bore depths are also not consistent, so you need to be careful which Acros parts you order, and they do a poor job documenting this.

Meanwhile. FSA, Token, Enve, Deda, and Cane Creek their designs to a relatively standard 8.7mm bore depth. I would just stick with one of those. Their designs work, and you can find the specs for them, and replacement parts are relatively easy to get.

[adam900710]

Hey, so I spend a lot of time with headset bearing since I run a microbusiness that helps people jailbreak from proprietary systems.

There are many misconceptions in this post. The most important consideration in headset design is how well it enables the bike to hold preload. The designs that are open in front are often better than those with routing on the sides for this. The contact area on the front is a minor part of the overall load path in these designs and is often offset by having a deeper C ring (see the FSA H2630A c ring or the Deda designs for example).

Then problem is not about pre-load, but when frontal crash happened, even minor ones. The C shaped one would easily eat into the steer tube.
AFAIK Luescher Teknik has too many videos on such damaged front forks.

Although I haven't had a frontal crash yet, this is not giving me any good impression.

The other thing to think about is how the hoses run. With the gap in front, the hoses can easily run down without needing extra stack height that takes them to the side. Different manufacturers handle this tradeoff differently. Pinarello is an extreme example where the routing path complexity is handled in the stem. Most do it like Specialized, Enve, and Acros, and do it in the spacer stack.

Since all I care is to go with separate bar and stem, and want enough routing headroom so that no hard hose bend, I don't think entering through the front brings any benefit at least to me.

And stack height is no big deal either, as I always have at least 20mm between the stem and top of the frame.

I don't argue with the premise that there is a lot of room for improvement, particularly with designs that are only made for 2 hoses, but I regretted playing around with the Acros system; it's unnecessarily complex and a PITA to get replacement parts for. The Acros design actually has an internal spacer, then a traditional metal compression ring, and then a headset cover. FSA, Token, Enve, Deda all use 2 pieces, just a headset cover and a compression ring. Compatible HT bore depths are also not consistent, so you need to be careful which Acros parts you order, and they do a poor job documenting this.

Meanwhile. FSA, Token, Enve, Deda, and Cane Creek their designs to a relatively standard 8.7mm bore depth. I would just stick with one of those. Their designs work, and you can find the specs for them, and replacement parts are relatively easy to get.

As long as the hoses go front in, it means either hard hose bend (Shimano Pro solution, or TOKEN A-BOX), for separate stem/bar setup, without really much headroom left.
Almost the same headroom as fully integrated routing.

Thus that's why I insist side entrance. BTW, the bigrock bearing seat has a depth of 8.4~8.5, not completely meeting the FSA 8.7mm requirement, thus it allows me to try ACROS solution. Will report back after they arrived.

I agree their parts are more available, but if your objective is to get something as close as the old semi-integrated cable routing with minimal hose bend and enough hose headroom so that you can remove the handle bar easily, just like the old days where hoses enter the head or down tube, I'd call they not a good middle ground.

Talking about availability, to my surprise, there are at least two domestic shops distributing ACROS headset, thus it's pretty good at least inside Australia.

[Sakizashi]

I think you still don't understand the function of preloading the steerer. The bending moment is greatly reduced by the preload (tension). The load path doesn't work as you are describing. C-shaped rings, if properly shaped, installed, and maintained, don't eat into carbon steerers. Having a gap in front or on the side also doesn't make a meaningful difference.

The big thing, and I think Raoul goes over this in his videos, is to ensure your headset is set up properly and maintained. The two major failures he points to in many cases are too many spacers (this can reduce the effective preload) and the headset coming loose (and therefore not being preloaded)

I guess given your considerations, there isn't much reason not do the ACROS. Their solutions work, but I wouldn't go so far as to say it's a better solution for most bikes. I am not sure about the hose-bend argument either; having worked with these, the front-opening headsets tend to have gentler bends for internal routing, with less strain on the hoses. With 20mm of spacers or even 10mm of spacers with under-stem routing, there won't be much, if any, difference.

[adam900710]

I'd say the best solution for me is to find a time machine and a timeline where no stupid fully integrated cable routing is ever invented.

42/52 is open and easily available, no hose bending problem at all.

The only problem is, I'm stuck in this timeline where even every chinese manufacture is pushing for this stupid idea.

[carbonMad]

I'd say the best solution for me is to find a time machine and a timeline where no stupid fully integrated cable routing is ever invented.

42/52 is open and easily available, no hose bending problem at all.

The only problem is, I'm stuck in this timeline where even every chinese manufacture is pushing for this stupid idea.

I think the best solution for you is to take a deep breath and obsess less over minute details.
Or, since you're running 'semi-integrated', just rotate the top cap and compression ring so they are slightly to the side instead of out front which is causing 'all your problems'.